Dec. 5, 1960 [CONTRIBUTIONFROX
INCORPORATION OF PHENYLALANINE INTO COLCHICINE THE
6055
SCHOOL OF CHEMISTRY, UNIVERSITY OF MINNESOTA, MINNEAPOLIS, MINNESOTA]
The Biogenesis of the Alkaloids of Colchicum. I. The Incorporation of Phenylalanine into Colchicine1 BY EDWARD LEETEAND PIROSKA E. NEMETH RECEIVED MAY9, 1960 The administration of ~ ~ - p h e n y l a l a n i n e - 3 t-oCsprouting ~~ Colchicum byzantinum corms led t o the formation of radioactive colchicine, Systematic degradation of the alkaloid showed that all the activity was located on one carbon strongly suggesting that phenylalanine is one of the precursors of colchicine. The results are discussed with respect t o previously suggested biogenetic schemes for colchicine.
The curative properties of Colchicum were known to the ancient civilizations of Greece, Egypt and India. Colchicine is the most important active principle found in the various species of Colchicum and this alkaloid has been the subject of extensive investigations by pharmacologists, biologists and chemists.2 Chemical investigations of colchicine have recently culminated in its total synthesis. This article is a n account of preliminary tracer work designed to elucidate its biogenesis in the plant. There have been several schemes proposed for the biogenesis of colchicine, and these will be briefly discussed since they have served as working hypotheses for our tracer studies. Anet4 suggested that colchicine may be biogenetically related to the flavones (111), the tropolone ring of colchicine being formed by the ring enlargement of the catechol ring by the insertion of a one carbon fragment.5 It has recently been established that phenylalanine is a precursor of the 3,4-dihydroxyphenylpropane moiety found in many flavonoid compounds, the other aromatic ring being derived from three acetate units6 Thus if the AnetRobinson scheme were correct the administration of phenylalanine-3-C14 (I) to Colchicum should yield colchicine (IV) labeled a t the position indicated in Fig. 1. Belleau’ suggested that colchicine is formed by the oxidative coupling of two molecules of 3,4,S-trihydroxyphenylpyruvic acid (V) to yield compound VI. The quinonoid ring of this compound then undergoes fission and cyclization of the resultant fragments shown diagrammatically in Fig. 2 results in the formation of the carbon skeleton of colchicine. Phenylalanine-3-C14 which (1) This work was presented in part a t the I.U.P.A.C. symposium on The Chemistry of Natural Products, Australia, August, 1960. This investigation was supported b y a research grant MY 2662, from the National Institute of Mental Health, Public Health Service, and by a grant from the research fund of the University of Minnesota. ( 2 ) cf. “Colchicine,” by 0. J. Eigsti and P. Dustin, The Iowa State College Press, Ames, Iowa, 1955. J. W. Cook and J. D. Loudon in “The Alkaloids,” Vol. 11, 1952, p. 261, and W. C. Wildman in “The Alkaloids,” Vol. VI, Ed. by R. H. F. Manske, Academic Press, Inc., New York, N. Y., 1960, p. 247. (3) J. Schreiber, W. Leimgruber, M. Pesaro, P. Schudel and A. Eschenmoser, Angew. Chem., 71. 637 (1959); E. E. van Tamelen, T. A. Spencer, D. S. Allen and R . L. Orvis, THISJOURNAL, 81, 6341 (1959). (4) F. A. L. Anet’s suggestion is recorded in the presidential address of Sir Robert Robinson to the Royal Society, A’ature, 166, 924 (1950). (5) This proposal for the mode of formation of some naturally occurring tropolones was first elaborated by Robinson in his presidential address.4 (6) T. A. Geissman and T. Swain, C h e m . and I n d . (London), 984 . , 227, 597 (1957); ibid., lab, (1957); H. Grisebach, 2. n a l u ~ f o ~ s c hlab, 335 (1958); H. Reznik and R. Urban, AJ~aturwissenschuffcn, 44, 592 (1957); E. W. Underhill, J. E. Watkm and A. C. Neish, Con. J . Biochem. Pkysiol., 35, 219, 229 (1957). (7) E. Belleau, Ezperienlia, 9 , 178 (1953).
?::a
HOOCJ”’
I/
\ /
/
I
HO
I
J
0
OH
I1
OH OH 111
IV
Fig. 1.-Anet-Robinson
scheme for colchicine biogenesis. 0
c-e-c
HO OH
V HO H O’
b H 0 vI Y
I
c,
OMe VI1
Fig. 2.-Belleau’s
C
/
C
scheme for colchicine biogenesis.
is a plausible precursor of V would thus be expected to yield colchicine labeled as shown in formula VII. Wenkerts proposed that colchicine is formed by the condensation of the protonated Schiff base VI11 and 5-hydroxytropolone (IX). He suggested that compound VI11 is formed from prephenic acid. However, i t is considered equally plausible that VI11 could arise from phenylalanine. If this were the case colchicine derived from I would be labeled as shown in formula X. Wenkert suggested that the tropolone I X was formed from shikimic acid. Phenylalanine is a metabolite of (8) E . Wmkert, ibid , 16, 165 (1959)
EDWARD LEETEAND PIROSKA E. NEMETH
6056
shikimic acid,s but i t seems unlikely that this biochemical transformation is reversible and we would not expect phenylalanine-3-C14 to yield colchicine labeled in the tropolone ring if this hypothesis is correct.
IX Fig. 3.-Wenkert’s
Vol. 82
activity in the radioactive colchicine was located on the carbon indicated by a heavy dot in formula X, Fig. 3. This decisive result is thus compatible with Wenkert’s hypothesis but does not agree with the schemes of Belleau and Anet as shown in Figs. 1 and 2. However, the possible relationship of colchicine to the biogenesis of flavonoid compounds is attractive and our results are consistent with a modified Anet-Robinson scheme which is illustrated in Fig. 4. In this scheme ring A of
X scheme for colchicine biogenesis.
Since all these biogenetic schemes involved phenylalanine or compounds closely related to it, it seemed logical to administer radioactive phenylalanine to Colchicum. Geiling’s group obOH OH tained radioactive colchicine by allowing C. autumXI XI1 naEe plants to grow in an atmosphere containing radioactive carbon dioxide.’O It was our original intention to use this species in our experiments since i t has received the most extensive investigation. However, our horticultural supplier was unaware of our scientific endeavors and provided us with C. byzantinum, a species native to Greece \ \ “ and Asia Minor, which has larger corms and proOH OH XI11 XI6’ duces more spectacular flowers. This species, like autumnale,belongs to the subgenus Eucolchicum and flowers in September. Fortunately this species proved to be a good source of colchicine. Phenylalanine-3-C14 was administered to the sprouting corms by a wick arrangement, details being recorded in the Experimental section. After one month the colchicine was isolated and purified until it had a constant specific activity. A further check on its radiochemical purity was obtained by acid hydrolysis to colchiceine” which had the same specific activity as the colchicine. The radioactive colchicine was oxidized with alkaline potas- colchicine corresponds to the 3,4-dihydroxyphenyl sium ferricyanide to yield 3,4,5-trimethoxyphthalic substituent of the flavones. The acetate derived anhydride using the same conditions as reported ring C in the intermediate X I reacts with formaldehyde or its biological equivalent to yield comby Haworth for the oxidation or’ chebulinic acid. This oxidizing agent gave better yields of the anhy- pound XII. Enlargement of ring C (perhaps facilidride than alkaline potassium permanganate which tated by prior phosphorylation of the primary had previously been used to oxidize colchicine.l 3 alcohol group) as indicated in formula XII, folThe anhydride was converted to 3,4,5-trimethoxy- lowed by condensation with ring A yields XI11 benzoic acid by refluxing with concentrated hy- which has the required carbon skeleton of colchidrochloric acid.I4 This acid was decarboxylated cine. Nuclear hydroxylation of the potential by heating in quinoline in the presence of copper tropolone ring yields XIV, which on elimination chromite catalyst, the evolved carbon dioxide being of water and loss of a hydroxy group affords comabsorbed in barium hydroxide to yield barium pound XV. Final steps are uneventful, involving carbonate. This barium carbonate and all the methylation of the four hydroxy groups, reductive other intermediates in this degradative sequence amination of the carbonyl group in ring B and acetyhad essentially the same specific activity as the lation of the resultant amino group. It has been colchicine (cj. Table I) indicating that all the recently shown16that six of the carbons of the tropolone ring of the mold metabolite puberulic acid (9) cj’. B. D. Davis in “Symposium on Amino Acid hfetabolism,” Ed. by W. M. McElroy and H . B. Glass, The John3 Hopkins Press, (XVI) are derived from acetic acid. This signifiBaltimore, Maryland, 1955, p. 799. cant result offers considerable support to our hypoth(lo) E. J. Walasrek, F. E . Kcisey and E . M. K . Geiliog, Science, 116, esis for the origin of the tropolone ring of colchi225 (1952). cine. We have fed acetate-Cl4 to Colchicum (11) S. Zeisel, Monatrh.. 7, 557 (1886). (12) R. D. Haaorth and L. B. de Silva, J . Ckcm. SOC.,3511 (1951). and our results will be reported in a later publi(13) A. Windaus, Sits. bcr. Heidelberg. Aknd. Wisr., 1 (1910). cation. (14) K. L. Alimchandani and A. N. Meldrum, J . Chcm. SOC.,117, 1164 (1920). There is a typographical error in this paper, the carboxyl
group in structures V and VI being in the wrong position.
(15) J. H. Richards and L. D. Ferretti, Biocktm. Biophyr. Rcrrarch Commxnr., No. 2, 107 (1060).
Dec. 5, 1960
INCORPORATION OF PHENYLALANINE INTO COLCHICINE
Administration of DtPhenylalanhe-3-C1' to Cotchicum byzantinum-Ker-Gawl and Isolation of the Colchicine.-It was not possible to inject the corms directly with a hypodermic syringe because of their hardness. A cotton thread was inserted into the surface layer (penetrating about 5 mm.) of a sprouting corm by means of a fine needle. The two ends of the thread were placed in a 5 ml. beaker containing a solution of the tracer. Ten corms were fed in this way with DLphenylalanine-3-Ca4 (17.4 mg. in 10 ml. of sterile distilled water, total activity-5.67 X 108 c.p.m.U). Each plant thus received 1ml. of solution which was absorbed in a few hours. Water was added daily to the corms by addition t o the beakers in which the ends of the cotton threads were placed. The sprouting corms produced pale purple flowers and remained healthy. After one month the corms and flowers were comminuted in a Waring Blendor with methanol. The mixture was filtered and the residue washed with more methanol to bring the total volume to 2.5 1. The air-dried marc weighed 260 g. The methanol solution was evaporated to dryness in vacuo and the residue dissolved in water (400 ml.), filtered and extracted with petroleum ether (b.p. 40SO') (2 X 200 ml.). The aqueous portion was made alkaline with sodium bicarbonate, filtered from a small amount of solid and then extracted with chloroform (7 X 200 ml.). The chloroform extract, dried over sodium sulfate, was evaporated in vacuo and the residue dissolved in benzene (15 ml.) and chromatographed on 60 g. of Woelm alumina (activity I).la The column was first eluted with 1:l benzenechloroform (100 ml.) and then with chloroform (500 ml.). The chloroform eluate was evaporated, redissolved in chloroform (15 ml.) and then diluted with petroleum ether (200 ml.) when a greyish white solid separated. Wution of this material in a small amount of chloroform followed by precipitation with petroleum ether was repeated three times yielding pure colchicine (235 mg.) which had an infrared spectrum identical with authentic colchicine.20 The radioactive colchicine obtained from the plant was diluted with inactive material and rechromatographed on alumina and crystallized from ethyl acetate until it had a constant (16) Melting points are corrected. Analyses were carried out by Mrs. Olga Hamerston and her assistants at the University of Minnesota. (17) Purchased from Research Specialties Co., Richmond, California. (18) Counts were carried out in a Nuclear-Chicago model D-47 Q gas flow counter using a "Micromil window." Determinations were carried out on samples of finite thickness, making corrections for efficiency and self absorption. (19) Purchased from Alupharm Chemicals, New Orleans, La. (20) T h e authors thank Dr. E. C. Rornfeld of the Eli Lilly Co. for a generous sample of colchicine. The alkaloid was also purchased from Nutritional Biochemicals Co.. Cleveland, Ohio, and purified by chromatography on alumina (6. J. N. Ashley and J. 0. Harris, J. Chcm. Soc., 677 (1944)).
6057
activity. The incorporation of tracer into the colchicine was 0.06%. Degradation of the Radioactive Colchicine. Colchicine .I1 -Colchicine (50 mg.) was dissolved in water ( 5 ml.) containing 0.03 ml. of concentrated hydrochloric acid and refluxed gently for 2 hr. The solution was filtered hot and on cooling pale yellow needles of colchiceine hydrate separated (29 mg.), m.p. 151-157'. Analysis on material dried a t 60" indicated a composition corresponding to the hemihydrate in agreement with Zeisel.ll Anal. Calcd. for G1HB0aN.0.5H20: C, 63.98; €I, 6.14; N, 3.55. Found: C, 64.20; H, 6.28; N, 3.58. 3,4,5-TrimethoxyphthalicAnhydride.-Colchicine (400 mg.) was added to a solution of potassium hydroxide (17.5 9.) and potassium ferricyanide (100 g.) in 400 ml. of water and heated on a steam-bath with stirring for 12 hr. At this time more potassium hydroxide (17.5 g.) and potassium ferricyanide (100 g.) were added. Similar additions of alkali and ferricyanide were made after 36 hr., and after 60 hr. the mixture was cooled and the yellow crystals of potassium ferrocyanide filtered off and washed with a little cold water. The filtrate was acidified with 50% sulfuric acid and the solution (400 ml.) extracted with ether in a continuous extractor for 24 hr. The ether extract was evaporated without drying and the residue sublimed (160-180°,~00.1 mm.). The pale yellow sublimate (76 mg., m.p. 144-140 )was crystallized from a mixture of bemetie and petroleum ether to yield colorless plates of 3,4,5trimethoxyphthalic anhydride (51 mg.), m.p. 146-147'. Anal. Calcd. for CIIH~~O':C, 56.46; H, 4.23. Found: C, 55.68; H, 4.45. 3,4,5-Trimethoxybenzoic Acid.-3,4,bTrimethoxyphthalic anhydride (41 mg.) was refluxed for 12 hr. with concentrated hydrochloric acid ( 1 ml.). The solution was filtered hot and on cooling colorless needles of 3,4,5trimethoxybenzoic acid (19 mg.) separated, m.p. 166-167', identical with authentic material (mixed m.p., infrared spectrum). Decarboxylation of 3,4,5-Trimethoxybenzoic Acid.-The trimethoxybenzoic acid (15 mg.) was refluxed with copper chromite catalyst (10 mg.) in quinoline (2 ml.) for 1 hr. in a current of carbon dioxide free nitrogen. The exit gases were bubbled through barium hydroxide solution yielding barium carbonate (11 mg.). The activities of the carrier free colchicine and its degradation products are recorded in Table I .
TABLE I All activities are in c.p.m./mM Specific act., x 10-1
Colchicine Colchiceine hemihydrate 3,4,5-Trimethoxyphthalicanhydride 3,4,5-Trimethoxybenzoic acid Barium carbonate
5.4 5.4 5.2 5.1 5.0